[grpc] Fix x86-windows build issue#17276
[grpc] Fix x86-windows build issue#17276ykadowak wants to merge 5 commits intomicrosoft:masterfrom ykadowak:master
Conversation
|
@yusuke-kadowaki ,Thanks for your contribution,please run the vcpkg x-add-version grpc --overwrite-version command,then commit again |
There was a problem hiding this comment.
@yusuke-kadowaki ,please add "port-version" : 1 in the vcpkg.json
|
The failures caused by popsift will be fixed by #17277. |
Wait, which vcpkg.json? I edited the grpc one from "port-version" : 3 to 1, which does not seem right. |
|
@yusuke-kadowaki ,sorry!I made a mistake, I posted its comment on you when I was reviewing another pr, my fault |
|
@JonLiu1993 Now I'm confused haha. I don't know why these tests are failing. Can you tell me what specifically you wanted me to do? |
|
You changed the portfile, so you shall increment the port version. Than commit this change. |
|
@dg0yt I was misunderstanding the versioning. Thank you for the comment. |
|
In order to avoid hard-coding so much about whether the target platform is executable, I've created an alternative PR #17424 which makes codegen a feature. You can then easily build it using @yusuke-kadowaki Could you try out that PR and let us know if it solves your issue? |
|
@ras0219 Anyway, |
|
The problem with that assumption is that arm64 is becoming much more prominent as a platform for all three OS's (Windows, OSX, and Linux). I think it's much better to have a consistent story that works for everyone, rather than trying to build a bunch of case-by-case stuff that is likely to need changes or be fully scrapped in the future. |
|
I understand that the hard-coding part is bad and needs fixed anyway. The only concern is, as I mentioned, it would be a bit confusing for x86-windows developer I think. What if you add something like this to you PR #17424. Just my opinion. I'm closing this and leave it to your PR. |
What does your PR fix?
Fixes Grpc installed libraries difference between x64-windows and x86-windows #17247
Does your PR follow the maintainer guide?
YesI think this is a regression of b5bb151.
The original PR contains a lot of changes but I only understand a small part of it. So I am not sure if this is a right way to fix this.